Human breast cancers can be classified into distinct subtypes that are tightly correlated to prognosis. The most common (~70%) are called luminal breast cancers that are positive for the receptors for the female hormones estrogen and progesterone. These tumors are hormone-dependent because they require estrogens to grow. Opposite are basal breast cancers (~15%), which are negative for steroid receptors, and are hormone- independent. Patients with luminal tumors have significantly higher disease free survival overall compared to those with basal tumors. Still, many luminal tumors become drug resistant, recur after years of remission, metastasize and cause death. One plausible explanation for this is the existence of a population of cells called tumor initiating cells that manifest tumor growth, drug resistance, and recurrence. Tumor initiating cells have been defined for clinical human breast cancers essentially by a single marker called CD44. We have identified a new putative marker of tumor initiating cells, cytokeratin 5, in luminal breast tumor models. The latter cells are negative for estrogen and progesterone receptors, and increase several fold in number with progestins, used in birth control and hormone therapy. The hypothesis tested in this proposal is that a subpopulation of cells in luminal human breast tumors that are steroid receptor negative, and cytokeratin 5 positive, represent drug resistant tumor initiating cells. Progestins expand this population of cells and would increase cancer risk. Alternative models in which cytokeratin 5 positive cells are selected and expand during and post treatments will be tested. Experiments in this proposal will test i) whether cytokeratin 5 positive cells in luminal breast tumors can reconstitute new luminal estrogen receptor positive tumors when placed in small numbers in vivo, ii) whether cytokeratin 5 positive cells preferentially survive treatment with conventional therapies directed at estrogen receptors and cell proliferation, iii) that progestins and other growth factors regulate the number of cytokeratin 5 positive cells in tumors, and iv) will identify drugs more effective at depleting this subpopulation of cells in luminal tumors. Completion of these studies will provide a prospective mechanism for how long term progestin use increases the pool of tumor forming cells, and subsequently an increased risk of tumor formation. Many women continue to take progestins for hormone therapy, birth control, or other health reasons, and an understanding of the risk involved is needed. This proposal will describe for the first time tumor initiating cells specific for luminal, hormone-dependent breast cancers. First line therapies for luminal tumors usually target the estrogen receptor positive cells; the estrogen receptor negative tumor initiating cells would likely escape these treatments, as well as chemotherapies, and will require destruction by new drugs.